AMD Piledriver rumours ... and expert conjecture

Page 272 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.
We have had several requests for a sticky on AMD's yet to be released Piledriver architecture ... so here it is.

I want to make a few things clear though.

Post a question relevant to the topic, or information about the topic, or it will be deleted.

Post any negative personal comments about another user ... and they will be deleted.

Post flame baiting comments about the blue, red and green team and they will be deleted.

Enjoy ...
 

Dude, like. dude.....i don't even....
picard-facepalm.jpg
 


If ARM proves to hold its own in the Desktop area and Server market, there won't be such an issue, I'm sure.



For the most part, AMD has done good things among the big mistakes in the short period of time after Phenom II came out.

Brazos is quite an amazing feat and the whole APU movement after it; I'm sure Intel would love to have that expertise (and patents). Also, we all might criticize BD's failed out come, but design wise, it's pretty good. Intel could take it and put some of their own and have a wonderful Server CPU design for a couple of gens.

And lastly, HSA. That thing is moving slowly, but so far it seems to be at an steady growing pace. Software that uses OCL is, little by little, starting to show tangible benefits, so it's a matter of milking it some money to get more dev on board with it. Intel has plenty of cash to do so if they do buy AMD indeed.

I really don't know if they'd buy the CPU division only, cause the GPU division sure is worth the money for Intel; we all know that.

Cheers!
 
if it's true, this looks like another case poor execution and aftermath of the previous poor execution. for example, remember how intel hyped quick sync - 'secret weapon', 'fastest fixed function encoder' and stuff? compare that to the amount of attention and effort amd gave to it's own fixed function encoder - vce, especially in trinity. vce support was broken for months and it made a bad first impression.
how is intel able to take market share in the low end if amd has better offerings? yeah, zambezi and desktop trinity perform worse at gaming with discreet cards, but they perform much better at other applications that can utilize higher number of integer units and trinity is very good for it's igpu. additionally, one can hide gaming cpu weakness by introducing weaker gpu like entry level to midrange cards. low end and cheap stuff is amd's niche, they shoulda done better.
btw, why does it say that intel owns 80%? i read somewhere else that amd owns 40% of the consumer cpu marketshare.... weird.
looks like dt llano chips remained unsold after all. not good.
the publicity on steamroller apus and improvement on cpu cores seems to have influenced people. amd's pr and amd-fanboys-who-think-they're-highly-informed-about-inner-workings-of-amd keep hinting that the next one is better. just look at the current situation - dt trinity barely hit the market and steamroller and excavator are being continuously hyped.
as for mobile, i can't believe amd i.e. trinity didn't perform per their potential. there may be more to this.
 


Well Intel has a lot more $$ to spend on R&D as well as cooperating with the software devs, but IMHO I think "VCE" was more of a "me too" marketing brochure checkbox item with AMD than a serious effort.

As for AMD's execution, we should find out more during their conference call, but my guess is that smartphones and tablets took away from both AMD and Intel, given the weak worldwide economy. It takes a pretty long time to shift market focus so in fact both Intel and AMD were caught with their pants down a bit, IMO, plus I'd bet the China & India markets were down as well. However it seems a sure bet that AMD failed to make up any ground in server, which is where the highest margins are, and in fact probably had to lower prices in order to maintain their tiny 5.5% share in that market. So, Zambezi being designed as a server CPU first, and later shoe-horned into the DT market seems to have failed on both fronts..

What I find surprising is how late AMD's announcement is - I'd think their stockholders would have cause for legal action since AMD had to know about their Q3 performance long before now. IIRC they announced their lousy Q2 before the quarter actually ended, instead of well into their 'quiet' period. And Intel gave out their revised guidance about a month ago IIRC..

how is intel able to take market share in the low end if amd has better offerings? yeah, zambezi and desktop trinity perform worse at gaming with discreet cards, but they perform much better at other applications that can utilize higher number of integer units and trinity is very good for it's igpu. additionally, one can hide gaming cpu weakness by introducing weaker gpu like entry level to midrange cards. low end and cheap stuff is amd's niche, they shoulda done better.

Intel is apparently making and selling a lot of Ivy's, won't know for sure until their Q3 results are out. They have already EOL'd the Sandy 2500K CPU in favor of Ivy. But just as AMD says their CPU is "good enough", Intel's HD4K iGPU may be "good enough" for Joe Blow when he shops at Walmart 😛.. Certainly browsing the web & checking email and even watching pr0n 😀..

btw, why does it say that intel owns 80%? i read somewhere else that amd owns 40% of the consumer cpu marketshare.... weird.

That was just the desktop segment, not the entire x86 market which includes server and mobile too. Maybe last winter AMD got up to 40% or higher desktop share, but I don't think so any longer. Historically DT has been their strongest area - around 25% - 35%.

looks like dt llano chips remained unsold after all. not good.
the publicity on steamroller apus and improvement on cpu cores seems to have influenced people. amd's pr and amd-fanboys-who-think-they're-highly-informed-about-inner-workings-of-amd keep hinting that the next one is better. just look at the current situation - dt trinity barely hit the market and steamroller and excavator are being continuously hyped. as for mobile, i can't believe amd i.e. trinity didn't perform per their potential. there may be more to this.

Yep, the ol' Osborne effect. Unfortunately when your next-gen products are late to the game, you have to start talking them up at some point to prevent marketshare erosion to the competition who does have their next-gen products (Ivy in this case) out and selling. If BD had come out when originally scheduled (December 2010) and so AMD engineers could devote all their attention to PD, they would have been in better market shape.

Of course, when talking up those next-gen products, you should strive for accuracy and not say things like IPC will be improved when it won't be 😛. Seeing as how that creates a bad impression when it actually does come out and the reviews are in..

 


I believe you know the answer, Zambezi is plenty to game and game well, in GPU applications there is parity to intel parts, in x86 bound applications it falls behind on average 15-20% but in terms of time or FPS generated its well and above necessary, when reviewed it will be told that it doesn't game well by rather calous reviewers. The problem has been reviews love to focus heavily on the negitives that the possitives are overlooked. Take Toms recent A10 vs i3 showdown, the A10 won most of the battles yet was given a ummm and a argh as to its worth at buying, yet in Game and Tech expo's the APU's draw the biggest attention, when people actually see what they do its a different kettle of fish. Benchmarketing whether you choose to accept it or not exists still today in the business and that will never change.

The other is exhibited in the attitude that people have basically everything to appear on AMD is met with disdain almost like its expected to be bad, this in stark contrast everything hyped about intel is almost agreed that it will be the case, intel has failed many more times than some care to give notice to. You can't have a chance if nobody takes notice.

Compared to the situation of a year ago AMD has done well to release a product on time, and with intended performance gains, they have also secured a new process and consolidated debts which obviously is not good for the short term balance books. As for layoffs, most are non engineering staff or non essential staff such is the case in global economic recessions, basically said that was going to happen just to bring it to green due to operational requirements.
 


I can see them buying the ATI portion, not the CPU portion at all.
 
Personally I'd be sacking Read ... his business decisions have been laughable.

The only results he has to glow about are as a direct result of initiatives his predessessor put into place.

His achievements to date are what ??

The guy hasn't even shown his face or posted a single comment since coming into office ... Hector II ...

All he has done is clean out the offices .. downsized everything.

Dirk at least had some sort of forward looking strategy.

I compare this guy to the current CEO of Qantas.

Both seem bent on breaking up the company ... so who will benefit?

Who is paying them to do it if that is the case ... hypothetically speaking?

 


The verdict is out as to motives, to go from one evil to another is not good for the company or investors alike. Its sad that Hector left after many years he should have been sacked and yet still walked out with a pretty fortune when I think a prison cell was more in order, unfortunately that is not a lonely cell either there are others that should be brought down too.
 


Give the guy some time. When he joined , AMD were in trouble. Its in worse trouble now financially, but atleast AMD has some good, some OK products in the market. Plus, the deal with SeaMicro is said to be a big thing for AMD.

One thing he is doing quite bad is losing engineeres to competitors. AMD needs good engineers.
 
^ Unfortunately GPU only accounts for a fraction of AMD's revenue.

And the news keeps getting worse: S&P may cut Advanced Micro Devices 'BB-' rating

Overview
-- U.S. semiconductor supplier Advanced Micro Devices Inc. (AMD) has
released preliminary financial results for the quarter ended September 2012,
which contained weaker results than we had anticipated and underperformance
relative to its primary competitor, Intel.
-- We now expect AMD's leverage to rise above our threshold for the
rating of 3.0x.
-- We are placing AMD's ratings, including its 'BB-' corporate credit
rating, on CreditWatch with negative implications.

Rating Action
On Oct. 12, 2012, Standard & Poor's Rating Services placed its ratings on
Sunnyvale, Calif.-based semiconductor microprocessor supplier Advanced Micro
Devices (AMD) Inc., including its 'BB-' corporate credit rating, on
CreditWatch with negative implications.

Rationale
The CreditWatch action follows the company's announcement yesterday that it
expects revenue for the September 2012 quarter to decline approximately 10%
sequentially, with weaker demand across all product lines, down from its prior
guidance range of -4% to 2%. The company also expects a 31% gross margin for
the September quarter, including the $100 million write-down of its inventory,
down from prior guidance of about 44%. We expect these results to contribute
to leverage remaining over 3.0x, up from 2.4x as of June 30, 2012, as well as
the potential that liquidity could approach the company's $1.5 billion cash
minimum target level, from about $1.8 billion as of June 30, 2012. In our
view, Intel's sequential revenue decline expectation for the September
quarter, which ranges between - 4.5% and flat sequential performance, implies
market share weakening for AMD during this period.

CreditWatch
We plan to meet with management to assess the company's business strategy and
financial profile before resolving the CreditWatch. We had previously
indicated that we could lower the rating if leverage were to stay above 3x or
if liquidity were to fall below $1 billion.

AMD down almost 14% today to $2.76. A lower credit rating means AMD will pay higher interest on any loans or notes they issue, obviously affecting future profits.
 


Well playing Russian Roulette and not blowing a hole in your head also works once or twice I guess 😀..

5 years ago AMD had a lot of assets to sell and credit worthiness they used to tide themselves over. Today, not so much.

 




Thats just a copyright for the logo. The date code look like:
FN 1229PGN

12 for 2012
29 for week 29 which puts it in July


You'll see the 2011 copyright on trinity A10/A8 pics as well.
 

5 years ago they were also at massive net loss, today they are at least making money, abide a small fraction of what a company their size should make.
 


Their market cap is half as much as it was in 2010. They don't actually make much money on GPU's vs CPU's. For some reason Amd keeps lowering its prices on GPU's when they don't need to. Even though they need more 7970 ghz editions at the right price.
 
With slowing sales in APU/CPU the last thing they want is slower GPU sales. Lowing the margins a bit can hedge off NVidia from gaining market share.

The last round of cuts was supposedly by the ODM/OEMs not AMD. The slump is causing fear of too much inventory.
 
AMD also doesn't have good execution in regards to market availability and supply timings.
Llanos were just starting wide availability 8 months ago and then trinity gets announced and delayed just 2 months later. Then trinity gets launched into a market over saturated with llano.

There is a rather large slump in the market and AMD is getting hit pretty hard. Investors are just panic selling for no reason as well since the general tone is bad for months and any small thing sets them off. Intel isn't doing that great either losing 30% of its stock value since the launch of ivy. Everybody is jumping ship because of the supposed post PC era thats suppose to come.
 
Status
Not open for further replies.